Cautious Evolution Underway in DTC Genetic Testing

Wendy K. Chung, MD, PhD

Order online. Spit in a cup. Learn how your genes affect your risk for hundreds of diseases including cancer.

When companies began offering such genetic testing directly to consumers more than a decade ago, they promised to spark a revolution in healthcare, not only by using new technology to make vital genetic information affordable but also by letting patients obtain that information without asking permission from doctors or insurers. Instead, they sparked controversy and drew scrutiny from regulators, which led the industry to abandon its original business plans. Now, changes are brewing again in the direct-to-consumer (DTC) genetic testing industry including developments with the potential to affect the public in the oncology field.

In February 2015, 23andMe became the first company to gain FDA clearance to market a DTC genetic test when the agency approved its saliva test for screening adults of reproductive age for Bloom syndrome, a rare disorder characterized by short stature, skin rash after sun exposure, and an increased risk of cancer.¹

In the fall, however, the FDA cracked down in another area—DTC pharmacogenomics testing. The agency sent letters to seven companies questioning recently introduced DTC genetic test products involving cancers and other diseases.²

Although one of the tests involved cancer hereditary risk assessment, the products included pharmacogenomics tests from four companies that purport to predict whether a person will respond to tamoxifen based on a genomic analysis of a cheek swab; one company’s testing also covers predictive responses to cyclophosphamide, docetaxel, and vincristine.

Over the years, the FDA’s oversight efforts of the DTC genetic testing industry have prompted debate over whether the agency has overstepped its authority and whether its actions would stifle innovation and ultimately harm patients. Others believe the current environment balances the need for innovation with the need to ensure that medical tests are accurate and useful.

Overall, observers say the consumer-focused testing industry of this new era is smaller and less ambitious than it once was. Only a handful of companies occupy the genetic risk prediction space, and none of them offer the sort of all-disease risk tests that were once common. It is, moreover, impossible to say how many risk prediction tests these companies perform because the privately held firms do not reveal sales figures. Regulators and researchers interviewed for this story wouldn’t even speculate about whether total test numbers have increased or decreased in the past few years.

In some quarters, the industry’s evolution is designed to assuage the concerns of its initial critics and stay on the right side of the government. Its leading companies use next-generation tests that they say virtually eliminate false negatives and false positives, and most require that doctors order tests for patients. They also tend to test only for genes with well-established risks that patients can actively address—a policy that leads to a heavy focus on the cancer-related mutations that have been the focus of so much recent research.

Initial Advances

“Direct-to-consumer testing businesses are continuously walking on a fine line, trying to find where the boundaries are,” said Jennifer Wagner, JD, PhD, associate director of Bioethics Research at Geisinger Health System, based primarily in Pennsylvania. “The FDA has tried to regulate DTC genetic tests by interpreting them as ‘medical devices,’ which is a quite a stretch, legally... From a technical standpoint, it has not forbidden DTC testing, but from a practical standpoint, it has scared many companies away from that business model.”DTC genetic tests first hit the market more than a decade ago as technologies to analyze the human genome advanced. They remained a niche item, however, until November of 2007, when the publicly traded company deCODE Genetics and the Google-backed startup 23andMe began selling tests. Navigenics followed soon afterward. These high-profile companies attracted widespread media coverage that introduced the “spit kit” concept to the American public and brought a few more competitors into the market.

Most companies offered a similar product. They charged fees that generally ranged, at first, from $1000 to $2000 to search each saliva sample for thousands of single nucleotide polymorphisms (SNPs) that had been linked (often tenuously) to some trait. Then, they used what they found to tell customers about everything from where their ancestors lived to what foods they should consider avoiding. Reports from the companies also told customers how their SNPs affected their risk for dozens or hundreds of different diseases.

Test providers attracted many enthusiastic supporters, who said that warnings about increased risk of diseases like cancer would save lives. They also attracted vocal critics, who complained of inaccurate results, insufficient data linking tested SNPs to disease risk, and many other perceived shortcomings. For example, the specificity of company reports often exceeded any possible justification. Using just a few SNPs—but no data on other known risk factors, including genetic factors that were too hard or too expensive to check—test companies would predict the exact risk of any patient developing a particular disease.

Another line of criticism attacked the very idea of allowing companies to provide complex and potentially devastating medical information directly to laymen without either regulatory oversight or professional mediation. The American Society of Clinical Oncology (ASCO) has tackled the sector several times in the past two decades, most recently in 2010, and cautions consumers about test reliability and interpretation concerns on its Cancer.Net website.³

>FDA Steps In

“The most fundamental principle of genetic testing that we feel should underlie genomic testing is the emphasis on the process of informed consent by a professional mediating the process so as to most responsibly insure the translation of this often complicated information to a widening range of individuals in the population,” said Kenneth Offit, MD, MPH, articulating ASCO’s position in a podcast.3 Offit is chief of the Clinical Genetics Service at Memorial Sloan Kettering Cancer Center.The FDA also saw problems with the DTC market that had evolved and decided to take action. In May 2010, the agency told Pathway Genomics Corp that its DTC tests appeared to meet the legal definition of a medical device and therefore required FDA approval before they could be sold. A month later, it sent similar letters to Knome Inc, Navigenics, deCODE Genetics, and 23andMe. “Marketing genetic tests directly to consumers can increase the risk of a test because a patient may make a decision that adversely affects their health, such as stopping or changing the dose of a medication or continuing an unhealthy lifestyle, without the intervention of a learned intermediary,” said Jeffrey Shuren, MD, director of the FDA’s Center for Devices and Radiological Health, speaking to a congressional committee that July.⁴ “The risk points up the importance of ensuring that consumers are also provided accurate, complete, and understandable information about the limitations of test results they are obtaining,” said Shuren.

The FDA did not forbid companies from selling genetic tests that predict disease risk directly to consumers. Instead, it said that companies would have to prove three essential elements to regulators before they could sell any such test:

1. The test methods accurately detect whatever genetic variations they seek with few false positives or false negatives.
2. The test seeks genetic variations that have well-established impact on a patient’s risk for developing a disease or passing it on.
3. The materials that explain the test and its results are easily understood by laymen.

“The agency, contrary to what many people seem to believe, wants companies to innovate and wants all people to have easy access to information about themselves,” FDA spokesman Eric Pahon said in an interview. “But it also wants to ensure that the information they do access is both accurate and intelligible. This is about protecting public health.”

Of all the companies that once provided health information via DTC genetic tests, only 23andMe has chosen to work with the FDA to continue a consumer-oriented model. Knome and deCODE left the consumer market and started selling to corporate and academic customers.

In September 2015, the FDA challenged Pathway Genomics over its CancerIntercept Detect test, which the company says can screen for up to 10 cancer types in high-risk populations through a blood sample that consumers collect themselves and ship to the company for analysis.⁵ The company said in a statement it would respond to the FDA’s concerns but that it believes the test is a laboratory- developed test appropriately marketed.⁶

Maintaining Consumer Appeal

Pathway currently is offering its tests through a hybrid between the DTC model and a traditional model designed to appeal solely to physicians; its tests are marketed to the public but must be ordered by a physician or healthcare provider.Other companies also require a doctor’s order before conducting any test. That said, they all try to appeal to patients in a way that traditional testing companies do not. They maintain consumerfriendly websites, price tests low enough for out-of-pocket purchase, report results in simple language, and employ genetic counselors to interpret results.

Among the companies that use this new model is San Francisco’s Counsyl Inc, a privately held firm that sells three tests: one that screens partners for genetic diseases they could pass on to children, one that looks for genetic ailments in embryos, and one that detects genes that increase cancer risk.

The company’s cancer test looks for whole genes rather than the SNPs that were the target of most first-generation tests. Whole-gene detection is very accurate. Counsyl said its testing platform produced few false positives or false negatives in a large study, and the company claims a 99.99% accuracy rate.

The Counsyl test also addresses another common complaint about older products—their tendency to report on hundreds or thousands of poorly understood mutations and leave patients without any clear course of action. Counsyl’s assay confines itself to 22 genes that not only have well-established ties to breast, ovarian, colon and/ or endometrial cancers but also, when detected, may affect medical recommendations for patient screening and care.

“When a test does come back positive for a mutation that increases cancer risk, it’s never just a scary piece of information that the patient can’t do anything about. It’s always something that justifies some type of action recommended by groups like the National Comprehensive Cancer Network to improve outcomes, and improving outcomes is the ultimate goal,” said Kaylene Ready, MS, CGC, director of Inherited Cancer Counseling at Counsyl. “In some cases, the recommendation may be preventative surgery or medication. In many cases, it’s additional screening. In cases where physicians order the tests after finding cancer, the results can help guide treatment. For example, genetic results that indicate how aggressive a cancer is likely to be can help breast cancer patients choose between a single small lumpectomy and a bilateral mastectomy.”

Utility of Testing

The cost of using traditional companies to test for so many genes would run to many thousands of dollars. Companies like Counsyl, Pathway, and Color Genomics that are more consumer-focused use propriety software and robotics to cut costs to several hundred dollars. The companies say these prices make it feasible for customers to finance their own tests when insurers won’t. They also say lower prices change the cost-benefit balance of genetic testing and justify far more thorough testing efforts.Research to date has not commonly found mutations for high-risk cancer genes in individuals who have cancer panel genetic testing if they do not have a personal or family history of cancer or are not of Jewish descent. The National Cancer Institute recommends individuals should consider genetic testing: (1) in the event of a personal or family history suggesting an inherited cancer risk; (2) if the test results can clearly show whether a specific genetic change is present or absent; and (3) the results provide information that can help guide treatment.⁷

That said, there are anecdotes of mutations in hereditary cancer genes that are incidentally identified, and it is hard to argue that outcomes would not improve if more people accurately knew their risk for cancer and the specific type of cancer.

“If testing costs fall far enough, then it certainly makes sense to test more patients. Not all of the individuals with high-risk genetic mutations that we know about meet current testing guidelines before they develop cancer. Identifying a larger percentage of them would allow us to improve outcomes for those people by tailoring cancer screening based upon their risk. There are specific medical actions that can be taken for almost all hereditary cancer predisposition,” said Wendy K. Chung, MD, PhD, the Kennedy Family Associate Professor of Pediatrics and Medicine at Columbia University.

Chung warns, however, that many cancers are not caused by genetic factors we can currently identify. That limits the ability of current genetic tests to predict all hereditary cancers and allow for enhanced cancer screening to enable early diagnosis. Most people who develop cancer do not have mutations in genes that are currently known to increase cancer risk. Our knowledge of which genetic variants increase cancer risk, moreover, is far from complete, even in the most thoroughly researched genes. It can be unclear, for example, whether an observed genetic variant in a cancer-related gene is pathogenic or not, particularly if the patient hails from an ethnic group that hasn’t been as well studied as Europeans.

The fact that test results can still be ambiguous can lead to misinterpretation by patients and, in some cases, increased anxiety. This was a consistent criticism of first-generation tests, a criticism that test providers have attempted to address by providing patients, along with doctors who don’t consider themselves subject experts, with free access to licensed genetic counselors. These counselors can explain the science behind the tests and the implications of their results. They can also direct people who receive disturbing news to groups that can provide a variety of support services.

“We have worked very hard to present results in the clearest possible terms, but we recognize, given the complexity of the subject, that many people will have questions, not only about what the results actually say, but also about what current guidelines say about what to do in response to test results. Our board-certified genetic counselors have many years of prior counseling experience answering those questions, not only for patients but also for the doctors who order the tests but may not consider themselves true experts in the subject matter,” said Elad Gil, PhD, who is CEO of Color Genomics, also a privately held company in the San Francisco area.

Color sells a $249 test that analyzes 19 genes associated with increased risk of breast or ovarian cancer, and it has been more aggressive than some of its competitors in reducing the barriers associated with ordering that test. The company will accept orders from any patient’s doctor, of course, but it also pays doctors to review information that patients submit online and either order the test for them or contact them with questions.

“Broadening access to this information is important. Only about 10% of the women who have the pathogenic mutations we test for have a strong enough family history of cancer to qualify for advanced screening. We’ve built a test that’s a tenth of the price of comparable genetic tests so that cost is not a barrier to accessing this information for physicians and their patients,” Gil said. “The key is broadening access in a responsible way, and we think the model we’ve developed does that.”

References

1. US Food and Drug Administration. FDA permits marketing of first direct-to-consumer genetic carrier test for Bloom syndrome. http:// goo.gl/3RSO5D. Posted February 19, 2015. Updated February 23, 2015. Accessed February 1, 2016.
2. US Food and Drug Administration. In vitro diagnostics news & events. http://goo.gl/Upx9G4. Updated December 9, 2015. Accessed February 1, 2016.
3. American Society of Clinical Oncology. Direct-to-consumer genetic testing. www.Cancer.Net website. Updated January 2014. Accessed February 1, 2016.
4. US Food and Drug Administration. Direct-to-consumer genetic testing and the consequences to the public. http://goo.gl/6KJfl3. Posted July 22, 2010. Accessed February 1, 2016.
5. US Food and Drug Administration. In vitro diagnostics news & events. Pathway Genomics letter. http://goo.gl/Upx9G4. Posted September 21, 2015. Accessed February 1, 2016.
6. Pathway Genomics’ response to FDA letter on CancerIntercept Detect [press release]. San Diego, CA; Pathway Genomics; September 24, 2015. Accessed February 1, 2016. https://goo.gl/XvVgUo
7. National Cancer Institute. Genetic testing for hereditary cancer syndromes. http://goo.gl/d8Vwfq. Updated April 11, 2013. Accessed February 1, 2016.